Content uploaded by Francois Cornet
Author content
All content in this area was uploaded by Francois Cornet on Mar 29, 2018
Content may be subject to copyright.
A preview of the PDF is not available
Analysis of Induced Seismicity for Stress Field Determination and Pore Pressure Mapping
Abstract
The focal mechanisms of some one hundred microseismic events induced by various water injections have been determined. Within the same depth interval, numerous stress measurements have been conducted with the HTPF method. When inverted simultaneously, the HTPF data and the focal plane solutions help determine the complete stress field in a fairly large volume of rock (about 15106 m3). These results demonstrate that hydraulically conductive fault zones are associated with local stress heterogeneities. Some of these stress heterogeneities correspond to local stress concentrations with principal stress magnitudes much larger than those of the regional stress field. They preclude the determination of the regional stress field from the sole inversion of focal mechanisms. In addition to determining the regional stress field, the integrated inversion of focal mechanisms and HTPF data help identify the fault plane for each for each of the focal mechanisms. These slip motions have been demonstrated to be consistent with Terzaghi''s effective stress principle and a Coulomb friction law with a friction coefficient ranging from 0.65 to 0.9. This has been used for mapping the pore pressure in the rock mass. This mapping shows that induced seismicity does not outline zones of high flow rate but only zones of high pore pressure. For one fault zone where no significant flow has been observed, the local pore pressure has been found to be larger than the regional minimum principal stress but no hydraulic fracturing has been detected there.
... In several instances, clear observations showed that in an hydraulically opened fracture system, the microseismicity was found to outline only zones with high interstitial pressure, but not helpful to identify the main flow zones away from the injection wells (Cornet and Scotti, 1993). Using a combination of HTPF measurements and induced seismicity focal mechanisms, Cornet and Yin (1995) managed to map the pore pressure in the rock mass, showing also that the induced seismicity did not outline zones of high-flow rates but only zones of high pore pressure. Those observations provided strong evidence that the location of the front of the microseismic cloud provided a good description of the motion of the pressure front. ...
... Third, it was found that passive seismic monitoring was blind to large changes related to the fluid flow experiment: the pressure increase induced shear and changes in the stress field, which were not detected due to aseismic behavior (Cornet and Scotti, 1993;Scotti and Cornet, 1994a,b). On the other hand, it is useful as an indicator of pore pressure change in the overall rock mass (Cornet and Yin, 1995). Fourth, arrays of surface tiltmeters could not be used to detect opening that should have been associated to fluid injection above the normal stress acting on some of the fractures that were taking fluid. ...
François Henri Cornet was Emeritus Professor at the Institut de Physique du Globe de Strasbourg, France and the author of the book "Elements of Crustal Geomechanics." Sadly, he passed away suddenly on May 23, 2020 in Strasbourg. He was well known to the ARMA community and was the MTS keynote speaker at the 2017 ARMA symposium in San Francisco. He was a PhD advisor, close colleague, and friend to many of us, and will always be remembered as a passionate leader and mentor with a big voice and laugh, always ready to engage in scientific discussions and a true expert on stress field evaluation. Among other things, he was the main inventor of the HTPF (Hydraulic Testing of Pre-existing Fractures) method for full stress tensor determination. At the time of his passing, he was working on his second book with co-authors called "Stress and stress fields in geomechanics." In this paper, we revisit François's main scientific contributions to the geomechanics and geophysics communities. His main research interests spanned many topics: hydraulic fracturing and in situ stress measurements, understanding and modeling of regional stress fields, induced and natural seismicity, rock-fluid interactions, development of new observations techniques, advocate of large scale natural geophysical laboratories (e.g., Le Mayet-de-Montagne, Soultz-sous-Forêts, Tongonan, Corinth Rift Laboratory, nuclear waste disposal and carbon storage sites), and new modeling techniques for volcanology.
... Commercial geothermal energy production requires a high geothermal gradient and is therefore often located in active tectonic regions (Brune & Thatcher, 2002). The size and rate of seismicity is then defined by the injection volume (and rate), the orientation of the tectonic stress field relative to the pore pressure increase and the extent of the deviatoric stress field within the local fault system (Cornet & Julien, 1989;Cornet & Jianmin, 1995;Brune & Thatcher, 2002). Grünthal (2014) analyzes the annual frequency-magnitude distribution of induced geothermal seismicity in central Europe in the period from 2000 to 2011 and compares it to the natural earthquake activity in the region. ...
The surveillance of geothermal seismicity is typically conducted using seismic networks, deployed around the power plants and subject to noise conditions in often highly urbanized areas. In contrast, seismic arrays can be situated at greater distances and allow monitoring of different power plants from one central location, less affected by noise interference. However, the effectiveness of arrays to monitor geothermal reservoirs is not well investigated and the increased distance to the source coincides with a decreased accuracy of the earthquake localizations. It is therefore essential to establish robust data processing and to obtain precise estimates of the location uncertainties. Here, we use time-domain array data processing and solve for the full 3-D slowness vector using robust linear regression. The approach implements a Biweight M-estimator, which yields stable parameter estimates and is well suited for real-time applications. We compare its performance to conventional least squares regression and frequency wavenumber analysis. Additionally, we implement a statistical approach based on changepoint analysis to automatically identify P- and S-wave arrivals within the recorded waveforms. The method can be seen as a simplification of autoregressive prediction. The estimated onsets facilitate reliable calculations of epicentral distances. We assess the performance of our methodology by comparison to network localizations for 77 induced earthquakes from the Landau and Insheim deep-geothermal reservoirs, situated in Rhineland-Palatinate, Germany. Our results demonstrate that we can differentiate earthquakes originating from both reservoirs and successfully localize the majority of events within the magnitude range of ML -0.2 to ML 1.3. The discrepancy between the two localization methods is mostly less than 1 km, which falls within the statistical errors. However, a few localizations deviate significantly, which can be attributed to poor observations during the winter of 2021/2022.
... Despite the seismic hazard posed by induced earthquakes (Foulger et al. 2018), these events are useful for assessing the development of the reservoir, as they can delineate the activated structures (e.g. Lengliné et al. 2017) in the reservoir, provide key indicators of the underground state of stress (Cornet & Jianmin 1995;Schoenball et al. 2014), track the evolution of the pore pressure (e.g. Shapiro et al. 2002), or help to infer geomechanical parameters of the reservoir (e.g. ...
Between November 2019 and July 2021, four induced earthquakes of local magnitude equal to or greater than than three were felt by the population of Strasbourg, France. These events were related to activity at the deep geothermal site GEOVEN located in Vendenheim in the northern suburb area of the city of Strasbourg. The first earthquake, with a local magnitude (Mlv) of 3.0, occurred on Nov 12, 2019, at the same depth as the bottom of the wells (approximately 4 km) but 5 km to the south. The second (Mlv 3.6) occurred a year later, on Dec 4, 2020, below the wells, and led to the termination of the project by the authorities. The third (Mlv 3.3) was initiated three weeks after shut-in on Jan 22, 2021, while the largest earthquake to date (Mlv 3.9) occurred on June 26, 2021, more than 6 months after shut-in. We constrained these four events’ absolute locations using a 3D velocity model of the area and here present regional intensity maps. We estimated moment magnitude and focal mechanism trough waveform inversion and inferred the fault plane activated during the largest event from an analysis of rupture directivity effects in the recorded waveforms. Our analysis highlights the existence of a critically stressed fault that hosted three of these widely felt events. We show how the derived source properties of these four earthquakes are directly linked to ground shaking observations at the surface. Notably, we demonstrate how earthquake moment, location, direction of rupture and stress drop impact the regional intensity distribution. Our results suggest that the traffic light system could benefit from including ground shaking scenarios based on realistic subsurface properties and potential earthquake source models.
... The model also ignores the effect of pore-pressure change on permeability. This is clearly an oversimplification as, in the case of fractured flow, the permeability increases substantially with pore pressure (Acosta & Violay, 2019;Cappa et al., 2006;Cornet & Jianmin, 1995;Evans et al., 2005). Common values of in-tact granite under comparable pressure are documented to be closer to 10 −21 m 2 (Brace et al., 1968), several orders of magnitude lower than that of the best fitting model (10 −16 m 2 ). ...
Induced seismicity observed during Enhanced Geothermal Stimulation at Otaniemi, Finland is modeled using both statistical and physical approaches. The physical model produces simulations closest to the observations when assuming rate‐and‐state friction for shear failure with diffusivity matching the pressure build‐up at the well‐head at onset of injections. Rate‐and‐state friction implies a time‐dependent earthquake nucleation process which is found to be essential in reproducing the spatial pattern of seismicity. This implies that permeability inferred from the expansion of the seismicity triggering front (Shapiro et al., 1997, https://doi.org/10.1111/j.1365-246x.1997.tb01215.x) can be biased. We suggest a heuristic method to account for this bias that is independent of the earthquake magnitude detection threshold. Our modeling suggests that the Omori law decay during injection shut‐ins results mainly from stress relaxation by pore pressure diffusion. During successive stimulations, seismicity should only be induced where the previous maximum of Coulomb stress changes is exceeded. This effect, commonly referred to as the Kaiser effect, is not clearly visible in the data from Otaniemi. The different injection locations at the various stimulation stages may have resulted in sufficiently different effective stress distributions that the effect was muted. We describe a statistical model whereby seismicity rate is estimated from convolution of the injection history with a kernel which approximates earthquake triggering by fluid diffusion. The statistical method has superior computational efficiency to the physical model and fits the observations as well as the physical model. This approach is applicable provided the Kaiser effect is not strong, as was the case in Otaniemi.
... Analysis of reopening pressures of fractures in different directions provides a method to constrain the entire stress tensor. The experimental results of the HTPF method have been presented by: Cornet and Burlet (1992); Cornet and Jianmin (1995), Cornet et al. (2003) and Wileveau et al. (2007). The shortcoming of the method is that it assumes negligible percolation of hydraulic fluid into the rock matrix, and that the stress state is homogeneous. ...
The influence of the intermediate principal stress, σ2, is disregarded when rock strength is analyzed using the Mohr-Coulomb or Hoek-Brown criteria. Several polyaxial strength criteria have been proposed to account for the influence of σ2. One of the methods that received considerable attention is the Mogi criterion. The criterion modifies the Nadai or Drucker-Prager approaches by exchanging the normal octahedral stress with what is introduced as the mean effective stress, σm,2. This change is based on experimental evidence that the influence of the intermediate principal stress on strength is less pronounced than the influence of the minimum principal stress. However, the Mogi criterion was derived somewhat empirically and consequently it is essential to analyze its potential theoretical deficiencies. In the current note, we analyzed the Mogi failure criterion, describing its characteristics in both τoct–σm,2 and σ1-σ2 spaces. We investigated the relation between Mohr-Coulomb and Mogi criteria, which marks the basis for the commonly used Mogi-Coulomb approach. We compared the performance of the Mogi criterion against available polyaxial strength data using three fitting approaches: Mogi-Coulomb based on triaxial compression results only, linear Mogi based on all polyaxial strength results, and power law Mogi also based on all data. The analysis reveals several serious deficiencies of the Mogi approach which all suggest that the Mogi criterion cannot properly represent the σ2 dependence of rock strength in many cases.
... The reservoir stimulation can be associated with local induced seismicity by affecting the present stress field state, which results in possible fault reactivation (Cornet and Jianmin, 1995;Tingay et al., 2005;Cornet et al., 2007;Bourouis and Cornet, 2009;Shapiro and Dinske, 2009;Moeck and Backers, 2011;Reis et al., 2013;Baouche et al., 2020a,b,c,d;Ganguli and Sen, 2020). We have modelled the fault reactivation potential at two units; the Silurian hot shales are exposed to an overpressure gradient (0.58 psi/ft) and high horizontal stress concentration. ...
This work aims to determine the present-day in-situ stress and pore pressure in the Ahnet Basin, Algeria, through a 1D geomechanical approach. We investigated the drilling-induced fracture (DIF) from FMI log data to ascertain the direction of maximum horizontal stress (SHmax) from the Ahnet Field. A mean orientation of N140° ( ± 10°) has been interpreted, which is NW-SE (N140°-N320°), with a local variation of ( ± 20°) compared to fields such as Illizi, Hassi Messaoud, and North Algeria, which can be explained by depth variation and intrinsic rock properties. A 1.05 psi/ft gradient of overburden stress (Sv) has been obtained from density. Pore pressure has been estimated from the sonic log by a normal compaction trend, indicates a hydrostatic regime from the surface to the top of the Silurian unit with an average pore pressure gradient of 0.43 psi/ft and an overpressure regime against the hot shale unit with a gradient of 0.58 psi/ft caused by the high in situ temperature in the study area and possible activity of the mega-shear zone. The poroelastic approach under transverse isotropic vertical conditions (VTI) has been used to calculate the minimum and maximum horizontal stress magnitudes. The outlines indicate a high-stress gradient close to 0.82 psi/ft, for Shmin calibrated with MDT stress points and 1.10 psi/ft for SHmax. The stress magnitudes results, suggest a present-day normal to strike-slip stress regime in the Ahnet Basin. Fault reactivation potential at two Silurian units has been inferred from the frictional theory analysis. The results indicate that increased pore pressure in hot shale formations due to by fluid injection and hydraulic fracturing causing shear slippage of the pre-existing faults, resulting in induced seismicity. Our study has contributed to the understanding of stress state origin in Ahnet Basin, the relationship between in situ temperature and pore pressure, and the fault stability analysis in such unconventional reservoir development.
The stress field controls patterns of crustal deformation, including which faults are likeliest to cause earthquakes or transmit fluids. Since the 1950s, maps of maximum horizontal stress (SHmax) orientations have advanced dramatically, and the style of faulting (relative principal stress magnitudes) has recently been mapped in some regions as well. This perspectives paper summarizes developments in characterizing stress orientations and (relative) magnitudes, including new seismic and borehole methods, as well as progress in identifying the causes of stress variations. Despite these advances, adding far more spatiotemporal detail would allow geoscientists to address many of today's key challenges regarding natural hazards, energy development, and geodynamics. In particular, it is critically important to characterize stress heterogeneity at multiple scales while also recognizing the coherent variability of the stress field. The second part of the paper considers how more detailed stress datasets could prove essential to addressing some of the grand questions in geoscience, including deciphering the poorly understood feedbacks between crustal dynamics and surface processes, improving earthquake and eruption forecasts, and determining the origins and shared properties of plate boundaries.
The necessity to reduce carbon emissions to mitigate climate change is accelerating the transition from fossil fuels to renewable energy sources. Specifically, hydropower, in particular, has emerged as a prominent and safe renewable energy source but entails reservoir-triggered seismicity (RTS). This phenomenon causes significant challenges for safe reservoir management. Irapé, in Brazil, is a prominent RTS site where seismicity surged after reservoir filling, with a maximum event of magnitude 3.0 in May 2006, just six months after the start of reservoir impoundment. Despite more than a decade has passed since the seismicity occurred, the factors governing these earthquakes and their connection to subsurface rock properties remain poorly understood. Here, we attempt to understand the potential causes of RTS at Irapé dam, which is the highest dam in Brazil with 208 m, and the second highest in South America. Permeability and porosity measurements of cylindrical cores from hard and intact rock samples, which have been extracted near the RTS zone, by pitting 10 cm from the surface, reveal a low-permeability rock. Porosity values range from 6.340 to 14.734 %. Only 3 out of the 11 tested samples present permeability higher than the lowest measurable value of the apparatus (0.002 mD), with the highest permeability being 0.0098 mD. The undrained response of the low-permeability rock placed below the reservoir results in an instantaneous increase in pore pressure and poroelastic stress changes due to elastic compression, which brings potential faults located below the reservoir closer to failure conditions. According to our analytical calculations, the increase in 136 m of the reservoir-water level caused a 0.54 MPa pore pressure buildup at the depth of the Magnitude 3.0 earthquake, i.e., 3.88 km, resulting in an increase of 0.82 MPa in the vertical effective stress and a decrease of 0.34 MPa in the horizontal effective stress. These changes resulted in an increase in the deviatoric stress that led to fault destabilization, causing the RTS. The laboratory measurements and analytical calculations corroborate the hypothesis that the initial seismic activity was induced by the undrained subsurface response to the reservoir loading at Irapé.
Cette thèse est consacrée à l’étude de la sismicité de la région Nord-Est de l’Algérie. Elle est la première du genre pour cette région, après l’installation du nouveau réseau sismologique Algérien.
Cette étude démontre de façon claire et définitive que cette région est la plus active du Nord de l’Algérie puisque près de 1/3 des évènements sismiques s’y produisent. D’autre part, on constate que cette sismicité ne touche pas de la même façon les différentes régions de cette partie de l’Algérie. C’est pour cela qu’à partir des données géologiques, tectoniques et sismologiques 9 zones sismogènes et plusieurs sous-zones ont été considérées. Par ailleurs, pour chaque zone et zones-zone sismogène, les valeur de b et Mmax ont été déterminés et comparés à d’autres études.
Dans les zones et sous zones sismogènes deux types de sismicité ont été mis en relief: une sismicité de type induite tel que l’exemple de la crise sismique de Mila en 2007 et une sismicité de type tectonique telle que la séquence sismique de Béni-Ilmane en 2010. Ces deux exemples de sismicité ont été révélées à partir d’études
approfondies suite à des campagnes d’enregistrements instrumentales d’études approfondies en termes de localisation et relocalisation, mécanismes aux foyers, champs de contraintes, paramètres de la source, CFF et la tomographie sismique.
Dans cette thèse, nous avons également mis en évidence de nouvelles failles actives (failles de Béni-Ilmane et la faille de Djebel Akhal) mais également considérées des failles peu étudiées précédemment (la structure transversale de Kherrata, la faille de Tachaouft, la structure d’El Madher et la structure de Bir-Haddada).
Cette thèse reste un document de référence pour plusieurs études ultérieures d’établir un modèle fiablede l’Aléa sismique dans ces régions.
Global efforts to tame CO2 emissions include the use of renewable energy sources, such as geo-energy harnessing. However, injecting pressurised fluids into the deep underground can induce earthquakes, hence converting CO2-related risk into seismic risk. Induced seismicity hazard is characterised by the overall seismic activity afb that is normalised by the injected fluid volume V and the parameter b of the Gutenberg–Richter law. The (afb,b) set has so far been estimated for a dozen of reservoir stimulations, while at least 53 geothermal fluid stimulations are known to exist, based on our survey. Here, we mined the induced seismicity literature and were able to increase the number of estimates to 39 after calculating afb from related published parameters and by imputing b with its expectation where this parameter was missing (0.65 ≤ b ≤ 2.9, with mean 1.16). Our approach was a two-step procedure: we first reviewed the entire literature to identify seismic hazard information gaps and then did a meta-analysis to fill those gaps. We find that the mean and median afb estimates slightly decrease from afb ≈ −2.2 to afb = −2.9 and −2.4, respectively, and that the range of observations expands from −4.2 ≤ afb ≤ 0.4 to −8.9 ≤ afb ≤ 0.4, based on a comprehensive review unbiased towards high-seismicity experiments. Correcting for potential ambiguities in published parameters could further expand the range of possibilities but keep the mean and the median relatively close to original estimates, with afb ≈ −2.3 and −2.4, respectively. In terms of the number of earthquakes induced (function of 10afb), our meta-analysis suggests that it is about half the number that could previously be inferred from published afb estimates (i.e., half the seismic hazard). These results are hampered by high uncertainties, demonstrating the need to re-analyse past earthquake catalogues to remove any ambiguity and to systematically compute afb in future geothermal projects to reduce uncertainty in induced seismicity hazard assessment. Such uncertainties are so far detrimental to the further development of the technology.
A method for determining the aspect ratio R and the principal directions of the regional stress field from a set of focal mechanisms is proposed. The method is validated first on an artificial set of data, then on a sequence of aftershocks observed after the 1971 San Fernando earthquake. It is then applied to the collection of focal mechanisms obtained from aftershocks observed after the November 23, 1980 Campania-Lucania earthquake of Southern Italy. Results suggest that the stress field in which aftershocks occurred is slightly different from that which prevailed before the main shock.-from Authors
Focal mechanisms of 31 Central France earthquakes (m(sub b) = 2.9-4.4) were evaluated to determine if slip is compatible with stress field determinations conducted at four nearby sites. The coherent picture that results indicates that the regional stress fields can be inferred by linear extrapolation of in-situ measurements performed between 400-1000 m depths. Furthermore, the results show that in spite of rather constant principal stress directions from one site to the next (100 km apart), principal stress amplitudes vary significantly. For the domain investigated (200x200 sq km), the size of the homogeneous stress field regions appears to be fairly small (about 100x100 sq km). In addition, within these regions, important local stress perturbations as identified by extremely inconsistent focal mechanisms, are present. Thus only by integrating different data types is it possible to map properly the regional stress field. This mapping provides means to identify the homogeneous regions as well as to highlight zones of local stress perturbations, most likely indicative of active faulting.
A triaxial acceleration-sense down-hole AE measurement system was developed and successfully applied to the monitoring of both man-made and natural hydraulic fracture in a geothermal field. The acceleration-sense instrumentation allows long-distance measurement of subsurface crack extension. AE events during hydraulic fracturing in Nigorikawa geothermal field could be detected at a distance of 1600m from the AE source. AE events during a closure operation of a well-head valve in the Kakkonda geothermal field could also be detected by the system. In this paper, the feasibility of the simple measurement system for in-situ monitoring of subsurface crack extension in a geothermal field is demonstrated with reference to the fracture mechanics model of a geothermal reservoir.
Hydraulic fractures may not develop perpendicularly to the least principal stress direction when the rock is anisotropic with respect to its strength. This has been taken to advantage to develop two heat exchangers in a shallow granite. First a hydraulic fracture was developed at 186 m in a first borehole; its orientation was defined by inspecting the borehole with a T.V.camera and using shut-in pressure measurement as well as a preliminary stress determination to show that the fracture had not turned away from the wellbore. A second well, bored 30 m away from the first one, intersected the fracture at the 156 m depth. The second heat exchanger was prepared by developing, from a third borehole, a second hydraulic fracture which intersected the first one somewhere between the two first boreholes.
The thermal and hydraulic charateristics of these heat exchangers have been investigated thru a 56 days circulation test for the first geometry and a 48 days test for the second one.
The mathematical model used for interpreting results from these two tests has been applied to the computation of cooling abacuses useful for the design of a deep hot dry rock heat exchanger.
The microseismicity induced by a fluid injection at a depth of 443 m in
a granitic rock mass has been monitored with a 16 station network. The
location of the five events observed on the near totality of the network
suggests that flow away from the well occurred along a preexisting
fracture nearly parallel to the maximum horizontal principal stress
direction and not along the fractures observed at the wellbore. Focal
mechanisms of these events are coherent with shear motions along
fissures which intersect the main flow path. A difficulty has been
encountered in the determination of source parameters from spectral
analysis of signals because of attenuation effects. Indeed, for
uncorrected spectra the corner frequency was found to decrease as the
source to station distance increased. This demonstrates the effect of
attenuation on the signals and therefore the necessity of correcting
spectra for attenuation effects before estimating source parameters.
The distributions of intensity and orientation of maximum shearing stress in typical stress systems are plotted on stereographic projections in order to show the three-dimensional relationships. Mathematical expressions of these relationships as well as graphical methods of evaluation are given. Relationships of shearing stress to orientation of fault planes and orientation of net slip of faults are suggested. Methods of studying the relationships of faults and shearing stress are described.
Several hundred microearthquakes with local magnitudes ranging between -4 and -2 were recorded during hydraulic stimulation experiments at the hot dry rock demonstration site, Fenton Hill, New Mexico. These events are probably caused by shear failure induced by high pore fluid pressures. Since the event locations seem to cluster in a narrow band near the hydrolic fracture, we were able to use microseismic techniques to locate the hydraulic fracture and monitor its growth. We calculate the minimum pore pressure increase necessary to induce rock failure using a simple model of slippage on preexisting fractures. By comparing a pore pressure distribution, calculated using a one-dimensional diffusion model with the distribution of event foci, we demonstrate that high pressures are a probable cause for these microearthquakes. Stress drops for these microearthquakes, calculated from spectral parameters, increased by a factor of 3 during the course of the experiment. The stress drop increase correlates with a migration of seismic activity away from the injection well bore, which suggests relatively low in situ shear stresses near the well bore caused by strain release in this region during previous pressurizations.